An unmanned aerial vehicle-based traffic anomaly detection method, device, equipment and medium
By acquiring obstacle information and dynamically adjusting the drone's shooting position, the problems of blind spots and obstructions in drone traffic anomaly detection have been solved, achieving more efficient traffic anomaly monitoring.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE THIRD RES INST OF MIN OF PUBLIC SECURITY
- Filing Date
- 2026-06-10
- Publication Date
- 2026-07-17
AI Technical Summary
In existing drone traffic anomaly detection, the fixed flight path leads to a fixed shooting position, resulting in blind spots for the onboard camera and obstruction by obstacles, leading to poor monitoring results.
By acquiring the geographical coordinates of abnormal events and obstacle information on traffic roads, the system plans the flight path of drones, dynamically adjusts the shooting position, collects a set of abnormal event features, and determines whether to adjust the drone position based on the completeness of the data, thus realizing a closed-loop iterative mechanism.
It effectively avoids interference from obstacles, reduces the risk of flight collisions, improves the accuracy and completeness of abnormal event detection, eliminates blind spots and obstructions, and enhances monitoring effectiveness.
Smart Images
Figure CN122416737A_ABST